Integrating Big Data into Program Evaluation: Results of a Supermarket-Based Nutrition Incentive Program in a Predominantly Latino Neighborhood

Introduction: Little is known about produce purchasing patterns among Latino SNAP recipients. In 2015, Latino populations were 1.2 times as likely as non-Hispanic whites to be obese, with the highest obesity rates in the U.S. of any ethnic group. Regular intake of fresh fruits and vegetables (FVs) can reduce the risk of obesity, stroke, cancer, type 2 diabetes, and all-cause mortality. Low-income individuals and families, including those receiving aid through the Supplemental Nutrition Assistance Program (SNAP), may be less likely to purchase FVs than cheaper, calorie-dense products. Ohio Produce Perks provides a monetary incentive for SNAP beneficiaries to purchase greater quantities of FVs. This incentive program was implemented in a large grocery store serving an urban neighborhood that is over 25% Latino, with nearly 50% of this Latino population receiving SNAP benefits.

Methods: Coupons were distributed to customers paying with a SNAP or Ohio Direction Card. Each coupon was worth $5 and could be redeemed for FVs. Transaction level data were uploaded through a secure File Transfer Protocol and stored on a protected server. For each transaction where a coupon was distributed or redeemed, a receipt was kept and entered into a database. Although we do not have demographic information about any specific coupon users, researchers collected the date and time of the transaction, SNAP dollars spent, number of items purchased, amount of SNAP remaining on the customer’s card, FVs purchased, and total amount spent on FVs from customer receipts. Researchers calculated days from coupon distribution to redemption. Paired t-tests were used to determine mean differences in SNAP spending, FV spending, and SNAP remaining at distribution and redemption.

Results: Of 328 coupons distributed, 217 were redeemed, averaging 2.4 redemptions per customer. On average, consumers took 15 days to redeem their $5 coupons. Much of the produce purchased included culturally-relevant foods within the Latino community, including yucca, plantains, and recao. FV spending at coupon redemption more than doubled from distribution, although overall SNAP spending remained similar.

Conclusion: Extensive supermarket receipt data allowed for examination of longitudinal changes in FV purchasing behaviors in a population of predominantly Latino SNAP participants. Coupon recipients did not spend more money during each shopping trip from coupon distribution to redemption, but did spend a greater proportion on FVs. This indicates that coupon incentive programs may help increase the proportion of FV spending among SNAP participants, which is particularly important in Latino populations.